The vertical dotted line represents the onset of 2010 Medicare fee cuts. The Truven Medicare and Commercial databases comprise large convenience samples of Medicare beneficiaries with Medicare Supplemental coverage for whom Medicare is the primary payer, as well as adults and families with commercial insurance from large US employers, respectively.
Zirui Song, Jacob Wallace, Hannah T. Neprash, Michael R. McKellar, Michael E. Chernew, J. Michael McWilliams. Medicare Fee Cuts and Cardiologist-Hospital Integration. JAMA Intern Med. 2015;175(7):1229–1231. doi:10.1001/jamainternmed.2015.2017
Physician practices are increasingly integrating with hospitals.1 For physicians, the expansion of accountable care organization contracts, centered on clinicians taking responsibility for population spending and quality, makes independent practice more challenging. For hospitals and health systems, acquiring practices helps them control referral patterns, coordinate care, and improve their bargaining power with payers.
In 2010, based on recommendations from the American Medical Association and a national practice expense survey of physicians, the Centers for Medicare & Medicaid Services reduced fees for cardiology services, focusing on those delivered in the office setting.2 For example, payment for a myocardial perfusion image in the office was cut 26%, compared with 5% in the hospital outpatient department (HOPD). Payment for an echocardiogram was cut 16% in the office, compared with a 3% increase in the HOPD setting. This widened the already existing payment gap favoring HOPDs—by 2013, an echocardiogram cost Medicare 141% more in HOPDs than in the office.3
The American College of Cardiology (ACC) projected a surge of integration in response to physician office fee reductions, with cardiologists exchanging practice ownership for more predictable salaries as hospital employees.4 We analyzed trends in cardiologist-hospital integration.
We analyzed 2007-2012 medical claims in a continuously enrolled national sample of traditional Medicare beneficiaries and commercially insured individuals from Truven Medicare and Commercial databases. We measured cardiologist-hospital integration by calculating the share of volume billed in HOPDs. This captures both shifts in care to HOPDs and changes in practice patterns induced by physician-hospital integration. We focused on 3 affected services—myocardial perfusion imaging (MPI), echocardiograms, and electrocardiograms.3 We expected shares of HOPD volume to increase.
We used segmented regression to assess changes in integration growth after the physician office fee cut. Independent variables included beneficiary age and sex, time trend, a postintervention indicator, and the interaction between postintervention and trend. We also included quarter and metropolitan statistical area fixed effects. Standard errors were clustered by metropolitan statistical area.
This research was approved by the Harvard Medical School Institutional Review Board.
Our sample included 806 266 Medicare beneficiaries with a mean age of 75.7 years, who were 53.3% female, and represented all states, and 12 567 069 commercially insured individuals aged 55 to 64 years who were 52.8% female with a similar geographic distribution.
Across all services, prices favored the HOPD setting after 2010 (Table). The shares of volume in the HOPD setting also increased after 2010 (Figure). Growth in the HOPD share was 5.9, 3.9, and 2.7 percentage points per year (P < .001) faster after 2010 compared with before 2010 for MPI, echocardiograms, and electrocardiograms, respectively. The overall volume of echocardiograms and electrocardiograms per beneficiary continued to increase after the fee cut, while that for MPI decreased slightly (Table).
Aggregate analyses of all cardiovascular imaging and cardiovascular medicine services produced qualitatively similar results. Similar results were also found in commercial populations, suggesting that integration was associated with comparable effects across payers (Table).
Integration accelerated after the fee cuts. This is consistent with the 2010 ACC Practice Census, which found that 40% of cardiologists planned to integrate with hospitals due to the fee cuts and 13% were considering it.5 The Medicare Payment Advisory Commission estimated that if cardiology imaging alone continued to migrate to HOPDs, nearly all would be provided there by 2021, costing an additional $1.1 billion per year to Medicare and $290 million per year in beneficiary cost sharing because of higher prices for facility-based services.3
Hospital outpatient departments may be more expensive than office settings because of the costs of licensing requirements, ancillary services, maintaining standby capacity, and treating more complex patients.3 However, if equivalent quality care could be delivered in the office, the case for paying the higher fee may be more difficult to justify. Moreover, while higher HOPD payments may be covering higher hospital costs, they may also be passed on to physicians through higher salaries. Ultimately, integration may offset savings that fee cuts were intended to achieve, both because facility-based fees are higher and because of higher prices due to market power.
Our results may not be causal or generalizable. Other market forces could have also encouraged integration, such as hospitals acquiring practices to preserve their referral base under new payment models and the rising costs of independent practice, including malpractice premiums, infrastructure costs (eg, electronic medical records), and costs of meeting new quality reporting or performance goals. Moreover, integration has not been limited to cardiology, supporting the potential effect of broader secular factors. At the service level, the effect of any fee cut depends on its magnitude, the previous fees in each setting, and changes in the volume of affected and substitute services across different sites of care.
Amidst growing recognition of payment disparities across sites of care, policies that aim to equalize payments across settings have received increasing attention. The president’s fiscal year 2016 budget proposes site-neutral payments, estimated to save nearly $29.5 billion over 10 years. If fee cuts did indeed lead to hospital acquisition of physician practices, then narrowing the payment gap may lead to less physician-hospital integration, which might in turn limit price increases from market power.6
Corresponding Author: Zirui Song, MD, PhD, Department of Medicine, Massachusetts General Hospital, 55 Fruit St, Boston, MA 02114 (email@example.com).
Published Online: May 26, 2015. doi:10.1001/jamainternmed.2015.2017.
Author Contributions: Dr Song had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Song, Wallace, McWilliams.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Song.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Song, Neprash, McKellar, Chernew.
Obtained funding: Song, McWilliams.
Administrative, technical, or material support: Wallace.
Study supervision: McWilliams.
Conflict of Interest Disclosures: None reported.
Funding/Support: Supported by grants from the National Institute on Aging (grant F30 AG039175) to Dr Song, from the Robert Wood Johnson Foundation/Health Care Financing Organization (grant 71408) to Drs McWilliams and Chernew, and from the National Science Foundation Graduate Research Fellowship (grant 1144152) to Mr Wallace.
Role of the Funder/Sponsor: The National Institute on Aging, the Robert Wood Johnson Foundation/Health Care Financing Organization, the National Science Foundation, and organizations with which the authors are affiliated had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; in the preparation, review, or approval of the manuscript; or in the decision to submit the manuscript for publication.